CN105335409A

CN105335409A - Target user determination method and device and network server

Info

Publication number: CN105335409A
Application number: CN201410373320.1A
Authority: CN
Inventors: 戴文渊; 何秀强; 曹国祥
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2014-07-31
Filing date: 2014-07-31
Publication date: 2016-02-17
Anticipated expiration: 2034-07-31
Also published as: EP3001332A4; EP3001332A1; WO2016015444A1; CN105335409B

Abstract

The invention discloses a target user determination method and device and a network server. The method comprises: for any target business, establishing correspondences between user marks of different users and behavior object marks of different behavior objects of the same type by obtaining historical data of the multiple behavior objects belonging to the same business type with the target business; based on the established correspondences, constructing a data model containing the user marks and the behavior object marks; by utilizing a numerical value update rule, calculating a probability value of the user corresponding to each user mark and becoming a target user of the target business; and by utilizing the probability value, selecting the target user of the target business. According to the target user determination method and device, a target user group can be openly determined, and the precision and efficiency of determining the target user are effectively improved; and the target user determination method and device have wide application ranges in practical applications.

Description

The defining method of a kind of targeted customer, equipment and the webserver

Technical field

The present invention relates to internet information processing technology field, particularly relate to the defining method of a kind of targeted customer, equipment and the webserver.

Background technology

Along with the development of Internet technology, internet plays more and more important effect in the life of people.People perform search by search engine, obtain the information oneself needed from internet, and the search key that Virtual network operator is then inputted by user assesses interest or the demand of user, is initiatively the product that user recommends to be associated with search key.

Such as: a is specially smart mobile phone that student designs, Virtual network operator is mated with the search key of product by the search key of user, the user needing this smart mobile phone can be determined, and the smart mobile phone designed for student is specially recommended user, with the product facilitating user's fast finding to need to oneself, and realize the effective popularization to product.

As can be seen here, how exactly for product determines the major issue that targeted customer becomes product and effectively promotes.

For large-scale search company, the search key (or user network page browsing record) generally inputted by user carries out mating the targeted customer determining commodity with commodity key word, but the not all search company of search key of user's input can both obtain, that is, the mode that the search key (or user network page browsing record) and the commodity key word that utilize user to input carry out mating the targeted customer determining commodity is only limitted to individual businesses and can uses, and determines that technical field can not be widely used targeted customer.

In addition, propose collaborative filtering, namely the customer group to this user with similar buying behavior is determined according to the buying behavior of user's history, and be this user's recommended products in conjunction with the buying behavior of customer group, this mode needs to analyze the buying behavior of user's history to mate with the buying behavior of other users, but once not have user to buy certain product (such as: new product), so this mode also will not use.

Summary of the invention

In view of this, the invention provides the defining method of a kind of targeted customer, equipment and the webserver, in order to solve in Products Show process how to determine targeted customer quickly and easily, to improve the precision and efficiency that targeted customer determines.

According to a first aspect of the invention, provide the defining method of a kind of targeted customer, comprising:

For arbitrary target service, the user behavior data that multiple object of action that acquisition and described target service belong to same type of service produce, wherein, contains user ID and object of action mark in each user behavior data;

According to the user ID comprised in the user behavior data obtained and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship;

According to the object of action comprised in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score;

Utilize described corresponding relation, build the data model being used for mark and transmitting, wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark;

Based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service.

In conjunction with in the embodiment that a first aspect of the present invention is possible, in the embodiment that the first is possible, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

Described based on described data model and described initial score, utilize numerical value update rule, interative computation is carried out to the numerical value of the element comprised in described data model, calculates the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, comprising:

According to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service.

In conjunction with in the first possible embodiment of a first aspect of the present invention, in the embodiment that the second is possible, described according to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, comprising:

Calculate the convergency value of the matrix element comprised in described transfer matrix; And

Determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

Wherein, the convergency value of the matrix element comprised in described transfer matrix is calculated in the following manner:

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

R (n) _mrepresent the convergency value of n the matrix element that the M time interative computation obtains, R (n) _m-1represent the convergency value of n the matrix element that the M-1 time interative computation obtains, α is decay factor, and T is transfer matrix, R (n) ₀contain the initial score of each user ID and the initial score of each object of action mark, n is natural number, represent in transfer matrix and contain n matrix element, value is the number sum that the number of user ID that comprises in the user behavior data obtained and object of action identify, m is natural number, represent the number of times performing interative computation, value is by the R (n) calculated _mwhether convergence is determined.

In conjunction with in the first possible embodiment of a first aspect of the present invention, or in conjunction with in the embodiment that the second of a first aspect of the present invention is possible, in the embodiment that the third is possible, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

For the user ID comprised in the user behavior data obtained, according to described corresponding relation, determine the number that the object of action that there is corresponding relation between a user ID identifies, and according to the number that described object of action identifies, obtain the initial value being identified the matrix element determined in transfer matrix by described user ID and the object of action that there is corresponding relation;

For the object of action mark comprised in the user behavior data obtained, according to described corresponding relation, determine and an object of action identify between there is the number of the user ID of corresponding relation, and according to the number of described user ID, obtain the initial value being identified the matrix element determined with the user ID that there is corresponding relation in transfer matrix by described object of action.

In conjunction with in the embodiment that a first aspect of the present invention is possible, or in conjunction with in the first possible embodiment of a first aspect of the present invention, or in conjunction with in the embodiment that the second of a first aspect of the present invention is possible, or in conjunction with in the third possible embodiment of a first aspect of the present invention, in the 4th kind of possible embodiment, the described object identity according to comprising in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score, comprise:

According to the object of action that target service comprises, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

In conjunction with in the embodiment that a first aspect of the present invention is possible, in the 5th kind of possible embodiment, determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, described method also comprises:

According to described corresponding relation, set up user ID and object of action identify between associated diagram, wherein, at least contain in described associated diagram user ID node, object of action identification nodes, possess incidence relation different user identification nodes between association line, the user ID that possesses incidence relation identify from object of action between association line, possess incidence relation different behavior object identity nodes between association line in one or more.

In conjunction with in the 5th kind of possible embodiment of a first aspect of the present invention, in the 6th kind of possible embodiment, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

Association line between identifying according to each user ID in described associated diagram and other user ID, object of action, determines to identify/or the initial value of matrix element determined of other user ID by described user ID and the object of action possessing incidence relation in transfer matrix;

According to the association line in described associated diagram between each object of action mark and user ID, determine the initial value being identified the matrix element determined with the user ID possessing incidence relation in transfer matrix by described object of action.

In conjunction with in the embodiment that a first aspect of the present invention is possible, or in conjunction with in the first possible embodiment of a first aspect of the present invention, or in conjunction with in the embodiment that the second of a first aspect of the present invention is possible, or in conjunction with in the third possible embodiment of a first aspect of the present invention, or in conjunction with in the 4th kind of possible embodiment of a first aspect of the present invention, or in conjunction with in the 5th kind of possible embodiment of a first aspect of the present invention, or in conjunction with in the 6th kind of possible embodiment of a first aspect of the present invention, in the 7th kind of possible embodiment, described user ID according to comprising in described user behavior data and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, comprise:

Determine the Social behaviors data of the user ID respective user comprised in the user behavior data obtained;

According to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation;

Utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

According to a second aspect of the invention, provide a kind of webserver, comprising:

Signal receiver, for for arbitrary target service, by the user behavior data that multiple object of action that communication network obtains and described target service belongs to same type of service produce, wherein, user ID and object of action mark is contained in each user behavior data;

Processor, for identifying according to the user ID comprised in the user behavior data obtained and object of action, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship; According to the object of action comprised in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score; Utilize described corresponding relation, build the data model being used for mark and transmitting, wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark; Based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service.

In conjunction with in the embodiment that a second aspect of the present invention is possible, in the embodiment that the first is possible, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

Described processor, specifically for according to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service.

In conjunction with in the first possible embodiment of a second aspect of the present invention, in the embodiment that the second is possible, described processor, specifically for calculating the convergency value of the matrix element comprised in described transfer matrix; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

In conjunction with in the first possible embodiment of a second aspect of the present invention, or in conjunction with in the embodiment that the second of a second aspect of the present invention is possible, in the embodiment that the third is possible, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

In conjunction with in the embodiment that a second aspect of the present invention is possible, or in conjunction with in the first possible embodiment of a second aspect of the present invention, or in conjunction with in the embodiment that the second of a second aspect of the present invention is possible, or in conjunction with in the third possible embodiment of a second aspect of the present invention, in the 4th kind of possible embodiment, described processor, specifically for the object of action comprised according to target service, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

In conjunction with in the embodiment that a second aspect of the present invention is possible, in the 5th kind of possible embodiment, described processor, also for determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, according to described corresponding relation, set up user ID and object of action identify between associated diagram, wherein, user ID node is at least contained in described associated diagram, object of action identification nodes, association line between the different user identification nodes possessing incidence relation, possess the user ID of incidence relation and object of action identify between association line, one or more in association line between the different behavior object identity nodes possessing incidence relation.

In conjunction with in the 5th kind of possible embodiment of a second aspect of the present invention, in the 6th kind of possible embodiment, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

In conjunction with in the embodiment that a second aspect of the present invention is possible, or in conjunction with in the first possible embodiment of a second aspect of the present invention, or in conjunction with in the embodiment that the second of a second aspect of the present invention is possible, or in conjunction with in the third possible embodiment of a second aspect of the present invention, or in conjunction with in the 4th kind of possible embodiment of a second aspect of the present invention, or in conjunction with in the 5th kind of possible embodiment of a second aspect of the present invention, or in conjunction with in the 6th kind of possible embodiment of a second aspect of the present invention, in the 7th kind of possible embodiment, described processor, specifically for determining the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition, according to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation, utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

According to a third aspect of the invention we, provide a kind of targeted customer locking equipment really, comprising:

Acquisition module, for for arbitrary target service, obtains the user behavior data that multiple object of action of belonging to same type of service with described target service produce, wherein, contains user ID and object of action identifies in each user behavior data;

Determination module, for the user ID that comprises in the user behavior data that obtains according to described acquisition module and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship;

Assignment module for according to the object of action comprised in described target service, for each user ID in described corresponding relation gives initial score, and is that in described corresponding relation, each object of action mark gives initial score;

Computing module, the described corresponding relation determined for utilizing described determination module, builds the data model being used for mark and transmitting, and wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark;

Based on the described initial score of described data model and described assignment module, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service.

In conjunction with in the embodiment that a third aspect of the present invention is possible, in the embodiment that the first is possible, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

Described computing module, specifically for according to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service.

In conjunction with in the first possible embodiment of a third aspect of the present invention, in the embodiment that the second is possible, described computing module, specifically for calculating the convergency value of the matrix element comprised in described transfer matrix; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

Wherein, the convergency value of the matrix element comprised in described transfer matrix is calculated in the following manner

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

In conjunction with in the first possible embodiment of a third aspect of the present invention, or in conjunction with in the embodiment that the second of a third aspect of the present invention is possible, in the embodiment that the third is possible, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

In conjunction with in the embodiment that a third aspect of the present invention is possible, or in conjunction with in the first possible embodiment of a third aspect of the present invention, or in conjunction with in the embodiment that the second of a third aspect of the present invention is possible, or in conjunction with in the third possible embodiment of a third aspect of the present invention, in the 4th kind of possible embodiment, described assignment module, specifically for the object of action comprised according to target service, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

In conjunction with in the embodiment that a third aspect of the present invention is possible, in the 5th kind of possible embodiment, describedly determine that equipment also comprises:

Associated diagram sets up module, for determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, according to the described corresponding relation that described determination module is determined, set up user ID and object of action identify between associated diagram, wherein, user ID node is at least contained in described associated diagram, object of action identification nodes, association line between the different user identification nodes possessing incidence relation, possess the user ID of incidence relation and object of action identify between association line, one or more in association line between the different behavior object identity nodes possessing incidence relation.

In conjunction with in the 5th kind of possible embodiment of a third aspect of the present invention, in the 6th kind of possible embodiment, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

In conjunction with in the embodiment that a third aspect of the present invention is possible, or in conjunction with in the first possible embodiment of a third aspect of the present invention, or in conjunction with in the embodiment that the second of a third aspect of the present invention is possible, or in conjunction with in the third possible embodiment of a third aspect of the present invention, or in conjunction with in the 4th kind of possible embodiment of a third aspect of the present invention, or in conjunction with in the 5th kind of possible embodiment of a third aspect of the present invention, or in conjunction with in the 6th kind of possible embodiment of a third aspect of the present invention, in the 7th kind of possible embodiment, described determination module, specifically for determining the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition, according to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation, utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

Beneficial effect of the present invention is as follows:

The embodiment of the present invention is for arbitrary target service, and the user behavior data that multiple object of action that acquisition and described target service belong to same type of service produce, wherein, contains user ID and object of action mark in each user behavior data, according to the user ID comprised in the user behavior data obtained and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship, according to the object identity comprised in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score, utilize described corresponding relation, build the data model being used for mark and transmitting, wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark, based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service, like this, for arbitrary target service, by obtaining the historical data belonging to multiple object of action of same type of service with this target service, set up the user ID of different user identify from the object of action of the different object of action of same type between corresponding relation, based on the multiple corresponding relations set up, build the data model comprising user ID and object of action mark, utilize iterative algorithm to obtain probable value that user corresponding to each user ID becomes the targeted customer of target service, and then utilize probable value to select the targeted customer of this target service, potential user group can not only be determined relatively openly, and effectively improve the precision and efficiency of determining targeted customer, in practical application, also usability is widely possessed.

Accompanying drawing explanation

In order to be illustrated more clearly in the technical scheme in the embodiment of the present invention, below the accompanying drawing used required in describing embodiment is briefly introduced, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.

The schematic flow sheet of the defining method of a kind of targeted customer that Fig. 1 provides for the embodiment of the present invention one;

Fig. 2 be set up user ID and object of action identify between associated diagram;

The schematic flow sheet of the defining method of a kind of targeted customer that Fig. 3 provides for the embodiment of the present invention two;

The structural representation of a kind of webserver that Fig. 4 provides for the embodiment of the present invention four;

Fig. 5 is the structural representation that the embodiment of the present invention five provides a kind of targeted customer locking equipment really.

Embodiment

In order to realize object of the present invention, embodiments provide the defining method of a kind of targeted customer, equipment and the webserver, for arbitrary target service, the user behavior data that multiple object of action that acquisition and described target service belong to same type of service produce, wherein, user ID and object of action mark is contained in each user behavior data; According to the user ID comprised in the user behavior data obtained and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship; According to the object identity comprised in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score; Utilize described corresponding relation, build the data model being used for mark and transmitting, wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark; Based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service.

Like this, for arbitrary target service, by obtaining the historical data belonging to multiple object of action of same type of service with this target service, set up the user ID of different user identify from the object of action of the different object of action of same type between corresponding relation, based on the multiple corresponding relations set up, build the data model comprising user ID and object of action mark, utilize iterative algorithm to obtain probable value that user corresponding to each user ID becomes the targeted customer of target service, and then utilize probable value to select the targeted customer of this target service, potential user group can not only be determined relatively openly, and effectively improve the precision and efficiency of determining targeted customer, in practical application, also usability is widely possessed.

Below in conjunction with Figure of description, each embodiment of the present invention is described in further detail.Obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making other embodiments all obtained under creative work prerequisite, belong to the scope of protection of the invention.

Embodiment one:

As shown in Figure 1, the schematic flow sheet of the defining method of a kind of targeted customer provided for the embodiment of the present invention one.Described method can be as described below.

Step 101: for arbitrary target service, the user behavior data that multiple object of action that acquisition and described target service belong to same type of service produce.

Wherein, user ID and object of action mark is contained in each user behavior data.

In a step 101, for arbitrary target business, before needing to determine the targeted customer of this target service, need the user behavior data (or historical data) that the multiple object of action obtained and this target service belongs to same type of service produce, based on the user behavior data produced, for this target service determines targeted customer.

Wherein, user behavior data refers to that user produces the data of a certain behavior in internet or real life for a certain object.

Object of action mark refers to the mark of the object that the operation of user's act of execution is corresponding, such as: application software, product information, Business Information, Word message, pictorial information etc.

Such as: user has downloaded an application software on the internet, and the user behavior data so produced is: comprise the user ID of this user and the download behavioral data of this application software mark;

User's first sets up the friend relation between user's second by instant communication software, the user behavior data so produced is: the Social behaviors data containing the instant messaging mark of user's first and the instant messaging mark of user's second;

User a have purchased product B at businessman A, and the user behavior data so produced is: the buying behavior data at least containing in the product identification of the user ID of user a, the merchant identification of businessman A, product B two kinds or three kinds; Contain the user ID of user a and the payment behavioral data of bank identifier;

User b is at M area call user c, and the user behavior data so produced is: the communication behavior data at least containing in the user ID of user b, the area identification in M region and the user ID of user c two kinds or three kinds; Etc..

Except user behavior data, also there are data that are a kind of and user-association, i.e. Social behaviors data.So-called Social behaviors data refer to the behavioral data comprising user social contact relation, such as: the friend relation set up between different user in immediate communication platform; Classmate's relation (such as: the classmate's relation set up in Renren Network platform) between different user; Etc..

That is, at current large data age, the behavioral data of people can record, by the analysis of the behavioral data to record, can set up incidence relation between user ID and user ID and/or user ID and object of action identify between incidence relation, foundation will be provided for the follow-up targeted customer of determination like this.

It should be noted that, object of action can comprise Word message, pictorial information, video information, can also comprise application software, client software, also can comprise Business Information, product information etc.

The mode obtaining user behavior data can be from the Website server of specifying, read the user behavior data recorded in a period of time periodically, also can be from the Website server of specifying, read user behavior data in real time, not limit here.

Due to user in internet to object of action executable operations time, the user behavior data that real time record user produces by server side in internet, such as:

Download behavior, server side will set up the corresponding relation between user ID and the application identities of download;

Buying behavior, server side by set up user ID, website logo and purchase product identification between corresponding relation;

Payment behavior, the corresponding relation etc. that server side will be set up between user ID, consumer products mark, account affiliated area mark.

When determining target service, according to the type of service of target service, determine the object of action belonging to same type of service with the type of service of this target service, and obtain from network data base and determine the user behavior data that object of action associates.

Such as: target service promotes 4G business of networking, can determine that target service belongs to data traffic types, the object of action belonging to same type of service with this target service type contains telephone expenses set meal, set meal etc. of surfing the Net, also just mean the telephone expenses package information, the online package information that need acquisition different user to use, therefrom find out the user likely using 4G business of networking.

Step 102: according to the user ID that comprises in the user behavior data obtained and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

Wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship.

In a step 102, according to the multiple user behavior datas obtained, analyze user ID in each user behavior data and object of action mark, select one of them object of action to identify, determine the user ID of different user and the object of action of selection identify between corresponding relation.

Such as: determine that this object of action of application software a identifies, from the user behavior data obtained, determine that user corresponding to which user ID performs operation to this application software a, the corresponding relation between the object of action of the user ID and this application software a of now setting up the user of executable operations identifies.

Again such as: the user behavior data obtained belongs to the user behavior data of downloading application software (such as: contain 4 user ID and 3 object of action marks in the user behavior data obtained, user ID is respectively U1 ~ U4, object of action mark is respectively A1 ~ A3), now, analyze the user behavior data obtained, for an object of action mark, determine the user ID of the application software that download behavior object identity is corresponding, and set up the corresponding relation between user ID and this object identity downloading application software corresponding to this object identity behavior, for a user ID, determine the object of action mark that application software that this user ID is downloaded is corresponding, and the corresponding relation between the object identity setting up the application software of this user ID and download, as shown in table 1:

User ID	Object of action identifies
		U1	A1
U1	A3
		U2	A2
U3	A2
		U4	A3

Table 1

As can be seen from Table 1, U1 has downloaded A1 and A3 respectively; U2 has downloaded A2, U3 and has downloaded A2, U4 and downloaded A3.

In another embodiment of the present invention, determine the user ID of different user identify from the object of action of the different object of action of described same type between the mode of corresponding relation include but not limited to under type:

First, the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition are determined.

Due to the incidence relation set up between user and user, a user can be made when using certain business, by the incidence relation between other users, business recommended to other users by what use, the incidence relation that is set up between user and user can determine potential targeted customer for promotion business.

Secondly, according to the Social behaviors data of the user ID comprised in the user behavior data obtained and the user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation.

Particularly, when obtaining user behavior data, with reference to the Social behaviors data of the user ID respective user comprised in this user behavior data, the direct correlation relation/between the different user mark determining to comprise in this user behavior data or indirect association relation.

Wherein, direct correlation relation refers to the direct Social behaviors data set up between the user that two different users mark is corresponding, such as: good friend each other between user A and user B; Or the communication behavior number of times between user A and user B exceedes setting threshold value, etc.

Indirect association relation refers to the direct Social behaviors data set up between the user of two different user mark correspondences and identical third party user, such as: good friend each other between user A and user C, good friend each other between user B and user C, so there is indirect association relation between user A and user B, etc.

Finally, utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

Still belong to the user behavior data of downloading application software (such as: in the user behavior data obtained, contain 4 user ID and 3 object of action marks with the user behavior data obtained, user ID is respectively U1 ~ U4, and object of action mark is respectively A1 ~ A3) be described for example.

When acquisition user behavior data, also obtain the Social behaviors data of U1 ~ U4, determine between U1 and U2, to there is direct correlation relation, there is direct correlation relation between U2 and U4, there is direct correlation relation between U1 and U3, as can be seen here, indirect association relation is there is between U1 and U4, there is indirect association relation between U2 and U3, between U3 and U4, there is indirect association relation, between U3 and U1, there is indirect association relation.

Determine so thus the user ID of different user and object of action identify between corresponding relation as shown in table 2:

Table 2

In another embodiment of the present invention, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation include but not limited to under type:

First, determine that the object of action comprised in the user behavior data of acquisition identifies the attribute data of corresponding object of action.

Wherein, whether the attribute data of object of action refers to whether object of action belongs to same developer, function identical or similar etc., such as: object of action identifies the different series that object of action that the object of action of 1 correspondence and object of action identify 2 correspondences belongs to same object of action, or, the function that the object of action of object of action mark 1 and object of action identify the object of action realization of 2 correspondences is identical, but does not belong to same developer etc.

Secondly, the object of action mark comprised in the described user behavior data according to acquisition and the attribute data of described object of action, determine the incidence relation between different behavior object identity.

Wherein, between the different behavior object identities with identical object of action attribute, there is incidence relation.

Finally, utilize the object of action mark comprised in the incidence relation between the different behavior object identities determined and described user behavior data, determine the user ID of different user and object of action identify between corresponding relation.

When acquisition user behavior data, also obtain the attribute data of the object of action of A1 ~ A3, determine to possess incidence relation between A1 and A2.

Determine so thus the user ID of different user and object of action identify between corresponding relation as shown in table 3:

User ID	Object of action identifies	The attribute data of object of action
			U1	A1	Incidence relation is possessed with A2
U1	A3
			U2	A2	Incidence relation is possessed with A1
U3	A2	Incidence relation is possessed with A1
			U4	A3

Table 3

It should be noted that, above-mentioned two kinds of being mentioned to determine the user ID of different user identify with object of action between the mode of corresponding relation can also be combined, do not limit here.

In another embodiment of the present invention, determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, described method also comprises:

Still for the content described in table 1, set up user ID and object of action identify between associated diagram, as shown in Figure 2.

Step 103: according to the object of action comprised in described target service for each user ID in described corresponding relation gives initial score, and is that in described corresponding relation, each object of action mark gives initial score.

In step 103, according to the object of action that target service comprises, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, to determine in described corresponding relation the initial score being greater than non-selected object of action mark by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with.

Wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

It should be noted that, for the user ID comprised in the user behavior data obtained and object of action mark, selectively can give initial score for a part of user ID (or object of action mark), other acquiescence initial scores not giving initial score are do not limit here.

Particularly, rule 1: for dissimilar object of action, the span of giving initial score is different.

Rule 2: the similarity identifying the object of action comprised in corresponding object of action and this target service according to the object of action comprised in the user behavior data obtained, determines the initial score of each user ID in the user behavior data obtained and the initial score of each object of action mark.

Wherein, the similarity of object of action refers to the similarity of object of action attribute, and such as: the functional similarity of object of action, or the function of object of action is identical etc.

It should be noted that, the similarity between the object of action that the value size of initial value can be promoted according to object of action corresponding to object of action mark and needing is determined, wherein, similarity is higher, and the value of initial value is larger.

Investigate object of action corresponding to object of action mark and need the similarity between the object of action promoted can be determined by similarity algorithm, also according to artificially to investigate or the mode such as experiment be determined, can not limit here.

It should be noted that, for each user ID in corresponding relation and each object of action mark, the initial score be endowed can be 0.

Step 104: utilize described corresponding relation, builds the data model being used for mark and transmitting.

Wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark.

At step 104, utilize described corresponding relation, build the data model being used for mark and transmitting, here data model can be the data model of a matrix form, it can also be the data model of a functional form, as long as about user ID and object of action identify and can carry out interative computation data model can, do not limit here.

Be that matrix is described in detail below with data model.

Described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark.

Namely the transfer matrix containing user ID and object of action mark is built.

Corresponding relation between still identifying for the object of action of the user ID shown in table 1 and downloading application software, the matrix element of the transfer matrix of structure put in order into: A1, A2, A3, U1, U2, U3 and U4, namely obtain the transfer matrix of 7*7.

Particularly, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

The original state of the transfer matrix now obtained is:

T = [\begin{matrix} 0 & 0 & 0 & 1 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & 1 / 2 & 1 / 2 & 0 \\ 0 & 0 & 0 & 1 / 2 & 0 & 0 & 1 / 2 \\ 1 / 2 & 0 & 1 / 2 & 0 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 1 & 0 & 0 & 0 & 0 \end{matrix}] .

Suppose, in a step 102, determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, also establish user ID and object of action identify between associated diagram, the determination mode of the initial value of the matrix element comprised in so described transfer matrix, can also comprise:

Step 105: based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service.

It should be noted that, numerical value update rule refers to a kind of iterative algorithm, by calculating the rule changing each element numerical value in transfer matrix, randomwalk class methods can be adopted, but the method for randomWalk class includes but not limited to Latticerandomwalk, Gaussianrandomwalk, and other variant form, do not limit here.

Particularly, the convergency value of the matrix element comprised in described transfer matrix is calculated; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID.

Particularly, the convergency value of the matrix element comprised in described transfer matrix is calculated in the following manner:

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

Wherein, R (n) _mrepresent the convergency value of n the matrix element that the M time interative computation obtains, R (n) _m-1represent the convergency value of n the matrix element that the M-1 time interative computation obtains, α is decay factor, and T is transfer matrix, R (n) ₀contain the initial score of each user ID and the initial score of each object of action mark, n is natural number, represent in transfer matrix and contain n matrix element, value is the number sum that the number of user ID that comprises in the user behavior data obtained and object of action identify, m is natural number, represent the number of times performing interative computation, value is by the R (n) calculated _mwhether convergence is determined.

Here, the R (n) of each user ID is judged _mthe mode whether restrained includes but not limited to, compares the R (n) that twice adjacent calculation obtains _mwith R (n) _m-1between difference, if difference is less than setting threshold value, then the R (n) of each user ID of calculating is described _mconvergence, if difference is not less than setting threshold value, then illustrates the R (n) of each user ID of calculating _mdo not restrain.

After obtaining the convergency value of each user ID, according to the size of convergency value, user ID sorted, the user selecting the larger some user ID of convergency value corresponding is as targeted customer.

Or, after obtaining the convergency value of each user ID, select convergency value to be greater than the user corresponding to user ID of setting threshold value as targeted customer.

It should be noted that, for the embodiment of the present application mention set up user ID and object of action identify between corresponding relation, in addition, the corresponding relation between user ID and user ID can also be set up, for setting up several groups of corresponding relations, incidence relation between the number of object that can relate to according to the user behavior data obtained and object is determined, is not specifically limited here.

Wherein, object at least contains one or more in user ID and object of action mark.

By the scheme of the embodiment of the present invention one, for arbitrary target service, the user behavior data that multiple object of action that acquisition and described target service belong to same type of service produce, wherein, contains user ID and object of action mark in each user behavior data, according to the user ID comprised in the user behavior data obtained and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship, according to the object identity comprised in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score, utilize described corresponding relation, build the data model being used for mark and transmitting, wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark, based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service, like this, for arbitrary target service, by obtaining the historical data belonging to multiple object of action of same type of service with this target service, set up the user ID of different user identify from the object of action of the different object of action of same type between corresponding relation, based on the multiple corresponding relations set up, build the data model comprising user ID and object of action mark, utilize iterative algorithm to obtain probable value that user corresponding to each user ID becomes the targeted customer of target service, and then utilize probable value to select the targeted customer of this target service, potential user group can not only be determined relatively openly, and effectively improve the precision and efficiency of determining targeted customer, in practical application, also usability is widely possessed.

Embodiment two:

As shown in Figure 3, the schematic flow sheet of the defining method of a kind of targeted customer provided for the embodiment of the present invention two.The embodiment of the present invention two thinks that application market determines that targeted customer is that example is described, and the method that this embodiment provides can be applied in online business datum field, such as: the fields such as electric business website, bank credit card business, commercial product recommending.

It should be noted that, the usual recording user in application market field browses the user behavior data such as address, down load application behavior.Such as: User downloads learning type application software, tourism user downloads and the user behavior data such as the application software of travel relevant (such as: contain tourist attractions information, Flight Information, hotel information).

Step 301: the application software used for a applicable student, obtain the user behavior data of application software type, and according to the application software mark containing user ID and user corresponding to user ID in described user behavior data and download, set up the user ID of different user and application software identify between corresponding relation.

The embodiment of step 301 is identical with the embodiment of step 101 ~ step 102 in the embodiment of the present invention one, is not specifically limited here.

Step 302: from the user behavior data obtained, to select and application software that the application software that uses of applicable student is identical or similar identifies, according to the application software mark selected, for each user ID in corresponding relation gives initial score, and each application software mark gives initial score.

In step 302, such as: be that to identify the initial score given be 1 to an application software identical with the application software that applicable student uses, giving initial score for other application software Identification is 0, and to give preliminary examination mark for the user ID in user behavior data be 0.

Step 303: utilize described corresponding relation, builds the transfer matrix being used for mark and transmitting.

Wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and application software mark.

In step 303, for the user ID comprised in the user behavior data obtained, according to described corresponding relation, determine the number that the application software that there is corresponding relation between a user ID identifies, and according to the number that described application software identifies, obtain the initial value being identified the matrix element determined in transfer matrix by described user ID and the application software that there is corresponding relation, such as: suppose that number is 2, the initial value so identifying by described user ID and the application software that there is corresponding relation the matrix element determined in transfer matrix is 1/2nd.

For the application software mark comprised in the user behavior data obtained, according to described corresponding relation, determine and an application software identify between there is the number of the user ID of corresponding relation, and according to the number of described user ID, obtain the initial value being identified the matrix element determined with the user ID that there is corresponding relation in transfer matrix by described application software, such as: suppose that number is 4, the initial value of the matrix element that the user ID so obtaining being identified by described application software in transfer matrix and there is corresponding relation is determined is 1/4th.

It should be noted that, according to the corresponding relation set up, do not set up the initial value that the user ID of corresponding relation or application software identify the matrix element determined and can be set as 0, do not limit here.

Particularly, build the transfer matrix comprising user ID and application software mark can be expressed as:

Wherein, for setting up the number of corresponding relation between u and a, u represents user ID, and a represents that application software identifies.

Suppose, the user ID comprised in the user behavior data obtained in step 301 is U1 ~ U4, and application software is designated A1 ~ A3, the corresponding relation between the user ID determined and application software identify:

Corresponding relation between U1 and A1, A2; U2 and A1, corresponding relation between A2, A3; Corresponding relation between U3 and A2, A3; Corresponding relation between U4 and A2, A3; Corresponding relation between A1 and U1, U2; A2 and U1, corresponding relation between U2, U3, U4; A1 and U2, corresponding relation between U3, U4.

Now, the transfer matrix obtained is:

T = [\begin{matrix} 0 & 0 & 0 & \frac{1}{2} & \frac{1}{2} & 0 & 0 \\ 0 & 0 & 0 & \frac{1}{4} & \frac{1}{4} & \frac{1}{4} & \frac{1}{4} \\ 0 & 0 & 0 & 0 & \frac{1}{3} & \frac{1}{3} & \frac{1}{3} \\ \frac{1}{2} & \frac{1}{2} & 0 & 0 & 0 & 0 & 0 \\ \frac{1}{3} & \frac{1}{3} & \frac{1}{3} & 0 & 0 & 0 & 0 \\ 0 & \frac{1}{2} & \frac{1}{2} & 0 & 0 & 0 & 0 \\ 0 & \frac{1}{2} & \frac{1}{2} & 0 & 0 & 0 & 0 \end{matrix}] .

It should be noted that, the application software mark that row matrix, column element are corresponding, the sequence of user ID are: A1, A2, A3, U1, U2, U3 and U4.

Step 304: utilize numerical value update rule and described initial value, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value of user corresponding to each user ID as targeted customer.

In step 304, the convergency value of the matrix element comprised in described transfer matrix is calculated; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID.

Particularly, the convergency value of the matrix element comprised in described transfer matrix is calculated:

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

R (n) _mrepresent the convergency value of n the matrix element that the M time interative computation obtains, R (n) _m-1represent the convergency value of n the matrix element that the M-1 time interative computation obtains, α is decay factor, and T is transfer matrix, R (n) ₀contain the initial score of each user ID and the initial score of each application software mark, n is natural number, represent in transfer matrix and contain n matrix element, value is the number sum that the number of user ID that comprises in the user behavior data obtained and application software identify, m is natural number, represent the number of times performing interative computation, value is by the R (n) calculated _mwhether convergence is determined.

Following iterative data is obtained through experiment:

R ¹⁰＝[0.1285，0.0786，0.0487，0.0713，0.0851，0.0305，0.0305]；

R ¹⁵＝[0.1311，0.0837，0.0526，0.0735，0.0884，0.0327，0.0327]；

R ²⁵＝[0.1317，0.0849，0.0535，0.0740，0.0892，0.0332，0.0332，0.0332]；

R ⁴⁴＝[0.1318，0.0850，0.0535，0.0741，0.0892，0.0332，0.0332]；

R ⁴⁵＝[0.1318，0.0850，0.0535，0.0741，0.0892，0.0332，0.0332]。

Now, by comparing R ⁴⁴and R ⁴⁵find, numerical value does not change substantially, now determines R ⁴⁵convergence, compares the numerical value that U1 ~ U4 is corresponding, and find that U2 is greater than U1, U1 is greater than U3, and U3 equals U4, now can determine that U2 is the targeted customer determined.

Embodiment three:

The embodiment of the present invention three is described in further detail with the scheme of telecom operators' data instance to the embodiment of the present invention.

It should be noted that, also there are the demand determining targeted customer in telecom operators, and such as, mobile phone towards certain user is promoted, tariff package recommendation etc.At the collectable user behavior data of telecom operators, comprising: the message registration of user, note record, set meal type, set meal expense.Simultaneously operator also makes a phone call according to user or eat dishes without rice or wine data logging in the base station of carrying out data access (during as online), infers the current location (architecture) of user.That is, for the business datum of telecom operators, the technical scheme that we also can use the present invention to propose carries out the determination of telecommunication service targeted customer.

First, utilize the mode of step 101 ~ step 102 in the embodiment of the present invention one build user ID, area identification, service package mark between corresponding relation.

Suppose, in the user behavior data of acquisition, contain U1 ~ U4 tetra-user ID, P1 ~ P3 tri-area identifications, F1 ~ F3 tri-service package marks, now, according to user behavior data, the corresponding relation between the user ID of foundation, area identification, service package identify:

The corresponding relation of the first type, the corresponding relation between user ID and user ID:

Corresponding relation between U1 and U3;

Corresponding relation between U3 and U4.

The corresponding relation of the second type, the corresponding relation between user ID and service package identify:

Corresponding relation between U1 and P1, P2;

Corresponding relation between U2 and P1, P2;

Corresponding relation between U3 and P2, P3;

Corresponding relation between U4 and P2, P3.

The corresponding relation of the third type, the corresponding relation between user ID and area identification:

Corresponding relation between U1 and F1;

Corresponding relation between U2 and F1;

Corresponding relation between U3 and F2;

Corresponding relation between U4 and F3.

Secondly, according to the corresponding relation set up, be respectively the user ID, area identification, the service package mark that comprise in this corresponding relation and determine an initial value, build the transfer matrix comprising user ID, area identification, service package mark.

Owing to there is multiple relationship type in this corresponding relation, we distinguish different relationship types.Such as, we sort to three kinds of corresponding relations according to the power of relationship type: " user ID-user ID ", " user represents-service package mark ", " user ID-area identification ".The ratio supposing the transmission effect for the present embodiment is 4:2:1, and the transfer matrix now obtained is: (in transfer matrix, order of elements is: F1, F2, F3, U1, U2, U3, U4, P1, P2, P3):

It should be noted that, according to set up corresponding relation, be respectively comprise in this corresponding relation user ID, area identification, service package mark determine that an initial value can be expressed as: I=[2/3,0,0,0,0,0,0,1/3,0,0], wherein, order of elements in initial value: F1, F2, F3, U1, U2, U3, U4, P1, P2, P3.

Finally, utilize numerical value update rule and described initial value, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculates the convergency value of each user ID, described convergency value is considered as the probable value of user corresponding to each user ID as targeted customer.

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

Wherein, the convergency value of n the matrix element that the M time interative computation obtains is represented, R (n) _m-1represent the convergency value of n the matrix element that the M-1 time interative computation obtains, α is decay factor, and T is transfer matrix, R (n) ₀contain the initial score of each user ID and the initial score of each object of action mark, n is natural number, represent in transfer matrix and contain n matrix element, value is the number sum that the number of user ID that comprises in the user behavior data obtained and object of action identify, m is natural number, represent the number of times performing interative computation, value is by the R (n) calculated _mwhether convergence is determined.

Embodiment four:

As shown in Figure 4, the structural representation of a kind of webserver provided for the embodiment of the present invention four.The described webserver possesses the function performing the embodiment of the present invention one to the embodiment of the present invention three, and the described webserver can adopt general-purpose computing system structure, and computer system can specifically based on the computing machine of processor.Described network server entity comprises signal receiver 41 and at least one processor 42, is connected between signal receiver 41 and at least one processor 42 by communication bus 43.

Processor 42 can be a general central processor (CPU), microprocessor, ASIC(Application Specific Integrated Circuit) (application-specificintegratedcircuit, ASIC), or one or more for controlling the integrated circuit that the present invention program's program performs.

Signal receiver 41, for for arbitrary target service, by the user behavior data that multiple object of action that communication network obtains and described target service belongs to same type of service produce, wherein, user ID and object of action mark is contained in each user behavior data;

Processor 42, for identifying according to the user ID comprised in the user behavior data obtained and object of action, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship; According to the object of action comprised in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score; Utilize described corresponding relation, build the data model being used for mark and transmitting, wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark; Based on described data model and described initial score, utilize numerical value update rule, the numerical value of the element comprised in described data model is calculated, obtain the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, and according to described probable value, select the targeted customer of described target service.

Particularly, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark.

Described processor 42, specifically for according to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service.

Particularly, described processor 42, specifically for calculating the convergency value of the matrix element comprised in described transfer matrix; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

Particularly, described processor 42, specifically for the object of action comprised according to target service, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

Alternatively, described processor 42, also for determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, according to described corresponding relation, set up user ID and object of action identify between associated diagram, wherein, user ID node is at least contained in described associated diagram, object of action identification nodes, association line between the different user identification nodes possessing incidence relation, possess the user ID of incidence relation and object of action identify between association line, one or more in association line between the different behavior object identity nodes possessing incidence relation.

Alternatively, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

Particularly, described processor 42, specifically for determining the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition; According to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation; Utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

Embodiment five:

As shown in Figure 5, for the embodiment of the present invention five provides the structural representation of a kind of targeted customer locking equipment really.Describedly determine that equipment comprises: acquisition module 51, determination module 52, assignment module 53 and computing module 54, wherein:

Acquisition module 51, for for arbitrary target service, obtains the user behavior data that multiple object of action of belonging to same type of service with described target service produce, wherein, contains user ID and object of action identifies in each user behavior data;

Determination module 52, for the user ID that comprises in the user behavior data that obtains according to described acquisition module and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, wherein, described corresponding relation is used for user corresponding to characterizing consumer mark and object of action and identify operation between corresponding object of action and by operative relationship;

Assignment module 53 for according to the object of action comprised in described target service, for each user ID in described corresponding relation gives initial score, and is that in described corresponding relation, each object of action mark gives initial score;

Computing module 54, the described corresponding relation determined for utilizing described determination module, builds the data model being used for mark and transmitting, and wherein, the element of the described data model of structure comprises user ID in described corresponding relation and object of action mark;

Particularly, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

Described computing module 54, specifically for according to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service.

Particularly, described computing module 54, specifically for calculating the convergency value of the matrix element comprised in described transfer matrix; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

Particularly, described assignment module 53, specifically for the object of action comprised according to target service, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

Alternatively, describedly determine that equipment also comprises: associated diagram sets up module 55, wherein:

Associated diagram sets up module 55, for determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, according to the described corresponding relation that described determination module is determined, set up user ID and object of action identify between associated diagram, wherein, user ID node is at least contained in described associated diagram, object of action identification nodes, association line between the different user identification nodes possessing incidence relation, possess the user ID of incidence relation and object of action identify between association line, one or more in association line between the different behavior object identity nodes possessing incidence relation.

Particularly, described determination module 52, specifically for determining the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition; According to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation; Utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

It will be understood by those skilled in the art that embodiments of the invention can be provided as method, device (equipment) or computer program.Therefore, the present invention can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the present invention can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) of computer usable program code.

The present invention describes with reference to according to the process flow diagram of the method for the embodiment of the present invention, device (equipment) and computer program and/or block scheme.Should understand can by the combination of the flow process in each flow process in computer program instructions realization flow figure and/or block scheme and/or square frame and process flow diagram and/or block scheme and/or square frame.These computer program instructions can being provided to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine, making the instruction performed by the processor of computing machine or other programmable data processing device produce device for realizing the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.

These computer program instructions also can be stored in can in the computer-readable memory that works in a specific way of vectoring computer or other programmable data processing device, the instruction making to be stored in this computer-readable memory produces the manufacture comprising command device, and this command device realizes the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.

These computer program instructions also can be loaded in computing machine or other programmable data processing device, make on computing machine or other programmable devices, to perform sequence of operations step to produce computer implemented process, thus the instruction performed on computing machine or other programmable devices is provided for the step realizing the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.

Although describe the preferred embodiments of the present invention, those skilled in the art once obtain the basic creative concept of cicada, then can make other change and amendment to these embodiments.So claims are intended to be interpreted as comprising preferred embodiment and falling into all changes and the amendment of the scope of the invention.

Obviously, those skilled in the art can carry out various change and modification to the present invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.

Claims

1. a targeted customer's defining method, is characterized in that, comprising:

2. the method for claim 1, is characterized in that, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

3. method as claimed in claim 2, it is characterized in that, described according to described initial score and numerical value update rule, interative computation is carried out to the numerical value of the matrix element comprised in described transfer matrix, calculate the convergency value of each user ID, described convergency value is considered as the probable value that user corresponding to each user ID becomes targeted customer corresponding to described target service, comprises:

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

4. method as claimed in claim 2 or claim 3, it is characterized in that, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

5. the method as described in as arbitrary in Claims 1-4, it is characterized in that, the described object identity according to comprising in described target service, for each user ID in described corresponding relation gives initial score, and be that in described corresponding relation, each object of action mark gives initial score, comprising:

6. the method for claim 1, is characterized in that, determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, described method also comprises:

7. method as claimed in claim 6, it is characterized in that, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

8. the method as described in as arbitrary in claim 1 to 7, it is characterized in that, described user ID according to comprising in described user behavior data and object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation, comprising:

9. a webserver, is characterized in that, comprising:

10. the webserver as claimed in claim 9, it is characterized in that, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

11. webservers as claimed in claim 10, is characterized in that,

Described processor, specifically for calculating the convergency value of the matrix element comprised in described transfer matrix; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

12. webservers as described in claim 10 or 11, it is characterized in that, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

13. as arbitrary in claim 9 to 12 as described in the webserver, it is characterized in that,

Described processor, specifically for the object of action comprised according to target service, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

14. webservers as claimed in claim 9, is characterized in that,

Described processor, also for determine the object of action of the user ID of different user from the different object of action of described same type identify between corresponding relation after, according to described corresponding relation, set up user ID and object of action identify between associated diagram, wherein, user ID node is at least contained in described associated diagram, object of action identification nodes, association line between the different user identification nodes possessing incidence relation, possess the user ID of incidence relation and object of action identify between association line, one or more in association line between the different behavior object identity nodes possessing incidence relation.

15. webservers as claimed in claim 14, it is characterized in that, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

16. as arbitrary in claim 9 to 15 as described in the webserver, it is characterized in that,

Described processor, specifically for determining the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition; According to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation; Utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.

17. 1 kinds of targeted customers locking equipment really, is characterized in that, comprising:

18. determine equipment as claimed in claim 17, it is characterized in that, described data model is transfer matrix, and wherein, the matrix element comprised in described transfer matrix contains user ID in described corresponding relation and object of action mark;

19. determine equipment as claimed in claim 18, it is characterized in that,

Described computing module, specifically for calculating the convergency value of the matrix element comprised in described transfer matrix; And determine the matrix element that each user ID is corresponding, using determine convergency value that matrix element is corresponding as calculate described matrix element corresponding the convergency value of user ID;

{R (n)}_{m} = α * T * {R (n)}_{m - 1} + \frac{1 - α}{2} * \frac{1}{n} + \frac{1 - α}{2} * {R (n)}_{0};

20. as described in claim 18 or 19 locking equipment really, it is characterized in that, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

21. as arbitrary in claim 17 to 20 as described in locking equipment really, it is characterized in that,

Described assignment module, specifically for the object of action comprised according to target service, to select from the user behavior data obtained and object of action that the object of action that comprises of described target service is identical or similar identifies, determine the initial score being greater than non-selected object of action mark in described corresponding relation by the initial score of the object of action mark selected, and determine to be greater than by the initial score that the object of action selected identifies the user ID that there is corresponding relation the initial score identifying the user ID that there is corresponding relation with non-selected object of action in described corresponding relation with, wherein, in described corresponding relation, identify with the object of action that described target service comprises initial score that identical object of action identifies and be greater than the initial score identifying similar object of action to the object of action that described target service comprises and identify.

22. determine equipment as claimed in claim 17, it is characterized in that, describedly determine that equipment also comprises:

23. determine equipment as claimed in claim 22, it is characterized in that, the determination mode of the initial value of the matrix element comprised in described transfer matrix comprises:

24. as arbitrary in claim 17 to 23 as described in locking equipment really, it is characterized in that,

Described determination module, specifically for determining the Social behaviors data of the user ID respective user comprised in the user behavior data of acquisition; According to the user ID that comprises in the user behavior data obtained and the Social behaviors data of user determined, the direct correlation relation/between the user ID setting up different user or indirect association relation; Utilize comprise in the user behavior data of the direct correlation relation/between the user ID of described different user or indirect association relation and acquisition object of action mark, determine the user ID of different user identify from the object of action of the different object of action of described same type between corresponding relation.